XJTU launches the "Jiaoxiaozhi" platform for teachers and students.

On April 30, Xi'an Jiaotong University (XJTU) unveiled the DeepEdu1.0 educational intelligent agent platform "Jiaoxiaozhi", which is jointly developed by the ‌National Key Laboratory of Human-Machine Hybrid Augmented Intelligence‌, the ‌Institute of Artificial Intelligence and Robotics‌, and the ‌Network Information Center‌.

The platform uses open-source models as its foundation for computational reasoning and content generation, while incorporating closed-source model API modules for specialized tasks. It employs natural human-machine interaction and combines intelligent plugins and modular components for research, teaching, and organizational management. The platform generates tailored content or code based on user requests, enabling teachers and students to create personalized AI agents.

The core human-machine interaction technology of "Jiaoxiaozhi" adopts the ‌hybrid-augmented intelligence framework‌ proposed in 2017 by ‌Academician Nanning Zheng‌ of the Chinese Academy of Engineering and a professor at XJTU.

This framework generates discipline-specific educational content, organizes fragmented teaching resources into structured knowledge graphs, creates generative teaching scenarios to enhance learning experiences, and provides personalized tutoring and interactive intelligent agents.

"Jiaoxiaozhi" offers multiple intelligent modules and plugins spanning diverse academic fields and scenarios. Teachers and students have built over ‌10,000 customized AI agents‌ for mathematics, humanities, engineering, clinical medicine, and other disciplines.

The platform supports smart lesson preparation, student profiling, learning strategy recommendations, academic intervention alerts, and other educational functions. It also integrates vertical-domain models, including ‌facial aging prediction‌, ‌medical image processing‌, and ‌smart agricultural seed selection‌.